Architecture

Technical overview of VideoCaptioner's architecture, processing pipeline, and development setup.

Processing Pipeline

Technical Highlights

• Word-level timestamps — Precise timing from Whisper with VAD refinement
• Semantic segmentation — LLM-driven sentence breaks based on meaning, not just timing
• Context-sensitive translation — Full subtitle context passed to LLM for coherent translation
• Concurrent batch processing — Multiple LLM requests in parallel for speed
• FFmpeg integration — Professional video synthesis with multiple output formats

Technology Stack

Development Setup

git clone https://github.com/WEIFENG2333/VideoCaptioner.git
cd VideoCaptioner

# Install with uv (recommended)
uv sync

# Run GUI
uv run videocaptioner

# Run CLI
uv run videocaptioner --help

# Type checking
uv run pyright

# Run tests
uv run pytest tests/test_cli/ -q

Project Stats

Contributing

Contributions are welcome! To get started:

1. Fork the repository on GitHub
2. Create a feature branch: git checkout -b feature/my-feature
3. Make your changes and write tests
4. Run uv run pyright and uv run pytest to verify
5. Submit a pull request

For bug reports and feature requests, use GitHub Issues.